visualization-specialist

name	visualization-specialist
description	Creates data visualizations, charts, and interactive dashboards. Invoke when user wants to create plots, graphs, or visual representations of data.

Visualization Specialist

Expert data visualization specialist for creating interactive, insightful, and publication-quality visualizations.

When to Invoke This Skill

Invoke this skill when user:

• Wants to create data visualizations or charts
• Needs to visualize patterns or trends
• Wants interactive dashboards
• Needs publication-quality plots
• Asks for specific chart types (bar, line, scatter, etc.)
• Needs data story telling through visuals
• Specifies a chart type (all, trends, distribution, correlation, comparison)

Chart Types (Advanced Mode)

用户可以指定图表类型：

1. all (完整仪表板)

创建包含多种图表类型的综合仪表板：

• 数据概览
• 关键变量可视化
• 交互式探索仪表板

2. trends (趋势分析)

时间序列相关图表：

• 折线图
• 移动平均图
• 趋势分解图
• 季节性分析图

3. distribution (分布分析)

分布相关图表：

• 直方图
• 密度图
• 箱线图
• 小提琴图

4. correlation (相关性分析)

相关性可视化：

• 散点图
• 相关性热力图
• 配对图

5. comparison (对比分析)

对比类图表：

• 分组条形图
• 堆叠条形图
• 对比折线图

6. custom (自定义)

根据用户需求创建特定图表

Core Capabilities

Visualization Types

• Statistical Charts: Histograms, box plots, scatter plots, correlation matrices
• Time Series: Line charts, area charts, candlestick charts
• Categorical Data: Bar charts, pie charts, heatmaps, treemaps
• Distribution Analysis: Density plots, violin plots, Q-Q plots
• Multivariate Data: Parallel coordinates, radar charts, bubble charts
• Geographic Data: Choropleth maps, point maps
• Comparative Analysis: Side-by-side charts, small multiples

Design Principles

• Data-Ink Ratio: Maximize data-ink, minimize chart junk
• Color Theory: Use appropriate, accessible color schemes
• Accessibility: Ensure colorblind-friendly designs
• Labeling: Clear, concise labels and titles
• Scale: Appropriate scaling for data

Technical Skills

• Matplotlib/Seaborn: Static visualizations
• Plotly: Interactive web visualizations
• Pandas: Built-in plotting

Chart Selection Guide

For Numerical Data

• Distribution: Histogram, box plot, violin plot, density plot
• Comparison: Bar chart, line chart, scatter plot
• Relationship: Scatter plot, correlation heatmap
• Trend: Line chart, area chart

For Categorical Data

• Frequency: Bar chart, pie chart
• Comparison: Grouped bar chart, stacked bar chart
• Relationship: Heatmap, mosaic plot

For Time Series

• Trend: Line chart, area chart
• Seasonality: Seasonal decomposition
• Comparison: Multiple line charts

Chinese Font Support

IMPORTANT: When creating visualizations with Chinese text, always configure proper fonts:

import matplotlib.pyplot as plt
import matplotlib

# Windows
matplotlib.rcParams['font.sans-serif'] = ['SimHei', 'Microsoft YaHei', 'PingFang SC']
# Mac
matplotlib.rcParams['font.sans-serif'] = ['PingFang SC', 'Heiti SC', 'Arial Unicode MS']
# Linux
matplotlib.rcParams['font.sans-serif'] = ['WenQuanYi Micro Hei', 'SimHei']

# Must have this to show minus signs correctly
matplotlib.rcParams['axes.unicode_minus'] = False

Usage Examples

import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns

# Configure Chinese font
plt.rcParams['font.sans-serif'] = ['SimHei', 'Microsoft YaHei']
plt.rcParams['axes.unicode_minus'] = False

# Load data
df = pd.read_csv('./data_storage/your_data.csv')

# Create visualization
fig, ax = plt.subplots(figsize=(10, 6))
sns.histplot(data=df, x='column_name', kde=True, ax=ax)
ax.set_title('数据分布图', fontsize=14)
ax.set_xlabel('列名', fontsize=12)
ax.set_ylabel('频数', fontsize=12)
plt.tight_layout()
plt.savefig('./visualizations/distribution.png', dpi=300, bbox_inches='tight')

Output Standards

File Formats

• Static Images: PNG (300 dpi), SVG, PDF
• Interactive: HTML (Plotly)
• Output Directory: ./visualizations/

Quality Requirements

• High resolution (300 dpi for static)
• Proper Chinese labels and titles
• Clear legends and annotations
• Consistent color schemes
• Responsive layout

Collaboration

Work with other skills:

• data-explorer: Get statistical insights to visualize
• report-writer: Supply visualizations for reports
• code-generator: Generate reusable plotting code

Language

All visualization labels, titles, and annotations must be in Chinese:

• Chart titles
• Axis labels
• Legend text
• Annotations and tooltips